Fix possible state reset issue

[S7evinK]

The spec says:

Iterative auth checks. The iterative auth checks algorithm takes as input an initial room state and a sorted list of state events, and constructs a new room state by iterating through the event list and applying the state event to the room state if the state event is allowed by the authorization rules. If the state event is not allowed by the authorization rules, then the event is ignored. If a (event_type, state_key) key that is required for checking the authorization rules is not present in the state, then the appropriate state event from the event’s auth_events is used if the auth event is not rejected.

The part we didn't do, given the membership event may not be resolved (e.g. it is conflicted):

If a (event_type, state_key) key that is required for checking the authorization rules is not present in the state, then the appropriate state event from the event’s auth_events is used if the auth event is not rejected.

[codecov]

Codecov Report

Attention: 2 lines in your changes are missing coverage. Please review.

Comparison is base (77f02e6) 58.76% compared to head (2c8d23d) 58.90%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #419      +/-   ##
==========================================
+ Coverage   58.76%   58.90%   +0.13%     
==========================================
  Files          51       51              
  Lines        7098     7105       +7     
==========================================
+ Hits         4171     4185      +14     
+ Misses       2536     2531       -5     
+ Partials      391      389       -2     

Files	Coverage Δ
stateresolutionv2.go	89.88% <80.00%> (-0.25%)	⬇️

... and 2 files with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[devonh]

Can you add a comment to the test outlining which prev_events are wrong?
It's impossible to tell what the event_ids are just from looking at the code.

Otherwise this looks like it should meet the rules outlined in the spec.

[S7evinK]

There isn't anything wrong with the prev_events in this test. The problem is mainly that Dendrite is "wrongly" calculating the conflicted events. See this.